Analysis of Snow Climatology Study

During a typical winter, many areas in the Upper Midwest experience the problem of blowing and drifting snow on roads and highways. The end result of the problem often includes a reduction in driver safety, degradation of road quality, and significant removal costs for drifted snow. Slick roads and poor visibility caused by blowing and drifting snow result in hazardous driving conditions. For example, from 1984–2001 there were 487 fatalities during the months of November–March due to snow, severe winds and blowing snow. Road quality can degrade due to the scraping of roadways and water infiltration induced by snowmelt. Furthermore, significant snow removal costs can be incurred by plowing and replowing windblown snow, not to mention secondary revenue impacts such as shipping delays, reduced commerce, lost salaries, and lost tax revenue.

The topography, land-use characteristics, and winter climate of western and southern Minnesota cause this area to be particularly vulnerable such that blowing and drifting snow is a common occurrence. The number of days with potential problems ranges from 115 in the south to 155 in the north. For an average winter season, taxpayers in Minnesota spend approximately $100 million in snow removal costs, with MnDOT expending $41 million. In the event of a winter season with anomalously high snowfall and exceedingly strong winds, as was the case for much of the state during the winter of 1996–97, the cost of snow removal can soar to $215 million.

Research conducted in the 1960s and '70s by the US Forest Service has shown that snow fences cause blowing snow to deposit on the landscape such that it is stored over the winter season. In the early 1970s, construction for Interstate 80 was completed in eastern Wyoming, however, the highway was subject to numerous drifting problems. Ronald Tabler and others utilized this as an opportunity to apply research results and deploy structural snow fences at problem locations along the interstate and test their effectiveness as snow control measures. It is estimated that these snow fences prevented 35 accidental injuries and cut winter maintenance costs in half for one season. Through these research efforts, detailed guidelines were established by Tabler (1994) outlining the steps necessary for the deployment of snow fences, both structural and living.

The Minnesota Department of Transportation chose to utilize the technique for site specific snow control and one study estimated that there are 4,000 sites, encompassing 1,000 miles, where blowing and drifting snow is problematic (Gullickson 1999). When comparing the cost of snow removal with living snow fences, an average cost/benefit ratio of 17:1 exemplifies the efficiency of this method validating widespread use in Minnesota. A research project was initiated by MnDOT and conducted in collaboration with the University of Minnesota's Department of Soil, Water, and Climate and the Minnesota State Climatology Office (DNR - Waters). Subsequently, a technical advisory panel was formed consisting of MnDOT, University of Minnesota, and USDA-NRCS affiliates.

Before a blowing snow problem can be solved, several climatological factors in the area of interest must be investigated. Climatological records of snowfall, precipitation, and temperature were compiled and archived in digital format for all observing stations in Minnesota. This database represents the climatic history of 370 locations throughout the state, some dating to the 1850s. This represents the most spatially and temporally comprehensive database available. In addition, wind speed and direction data from airport observing stations in Minnesota and surrounding areas were also compiled.

With these data in hand, the climatological parameters needed to design a living snow fence were computed. A knowledge of three attributes is required, length of the snow season, snowfall during that season, and the potential snow transport.

To define the first attribute, a method based on mean monthly temperature was used, as described by Tabler (1994). This allowed for the calculation of the beginning and ending dates of the snow accumulation season across Minnesota. Secondly, the mean and frequency distribution for snowfall during the snow season was determined. Finally, to calculate potential snow transport, a set of models was utilized. The snowfall and wind data, along with the precipitation data (used to determine snow water equivalent), allowed for calculating the potential snow transport

Since snow fences are most often deployed on land used for crop production, agricultural implications represent a chief concern for landowners. For the winter of 2000–01, three locations in southern Minnesota, representing three types of living snow fence designs, were studied. Each fence (previously planted) was intended to protect a known problem section of highway.

1. Two eight-row strips of corn (planted spring of 2000) placed 150 feet apart were left standing in a field north of Gaylord, Minnesota, in Sibley County to protect a portion of Trunk Highway 22. The landowner was compensated through MnDOT's standing corn row program at a rate of $2.75 per bushel for the estimated number of bushels of corn left standing in the field.
2. A twin-row 1998 planting of honeysuckle protects Highway 14 just west of Lamberton, Minnesota, in Redwood County. This fence was planted cooperatively by the University of Minnesota, MnDOT, Redwood Soil and Water Conservation District, and the Natural Resource Conservation Service.
3. North of Mountain Lake, Minnesota, in Cottonwood County, a section along Highway 30 proved to be a continuous problem for travelers as well as the nearby landowner that often came to the rescue for vehicles stuck in the snow. An alternating honeysuckle and red cedar single living snow fence was planted 300 feet from the north side of this highway in 1998.

Several features of the impacts of living snow fences were investigated. Snow storage was measured at the end of the snow season, which allowed for estimating the total volume of water that was caught and stored at each fence. Soil temperature and the depth to freezing was measured and studied in relation to snow depth. Soil moisture was sampled adjacent to each fence after snowmelt to investigate spatial variability, and finally a crop yield analysis was performed for areas surrounding each fence that were affected by increased snow deposition. The winter of 2000–01 proved to be an excellent season for this type of research because the total snowfall amount ranked in the highest 5 percent at these locations out of the last 100 years.

• The climatological database that was compiled represents the best spatial and temporal coverage possible for Minnesota.
• The user is able to access daily climatological records of temperature, precipitation and snowfall for all available stations in Minnesota back to 1891.

• During the 20th century, the annual snowfall in Minnesota has increased by an average of 1.5 inches per decade; therefore, it is advised that the mean snowfall for the most recent 30-year period (1971 –2000) be incorporated in the snow transport model when designing a living snow fence.
• Analysis of hourly wind observations showed that the predominant winter wind direction and the direction of greatest snow transport may not be consistent, and the latter should be utilized for snow fence design purposes.

• New wind-blown snow relocation coefficients have been calculated for Minnesota and are shown to vary geographically with higher values in western portions of the state. These values should be used in the mean seasonal snow transport model, which helps the user determine the appropriate fence design.
• The three living snow fence designs performed exceptionally well for the near record-setting winter of 2000–01, storing 50 percent of their theoretical capacity (based on an empirically developed equation using data from structural fences, Tabler 1994) and protecting their respective roadways from drifted snow.
• Given the high snow storage at each fence and record-setting rainfall during and after snowmelt, there was a relatively minor influence on crop yields adjacent to the snow fences.
• A 10-foot wide, black geotextile fabric ("EarthMat") was placed on top of the soil surface at Lamberton to help establish the honeysuckle plantings. Comparative soil temperature measurements at a two-inch depth beneath the fabric and bare soil showed that maximum daily temperatures in summer were lower under the fabric, suggesting high heat loads were not a factor hindering survival of new plantings.

• A user is now able to view previously unavailable snowfall and wind climatological products and utilize these products to follow the guidelines to design a living snow fence in an interactive, web-based environment.

Technical Advisory Panel (TAP) Members

University of Minnesota 

• Mark Seeley, Professor (Principal Investigator)
• Martha Shulski, Graduate Research Assistant
• Kenny Blumenfeld, Research Technician
• Amy Cecchi, Web Designer

Minnesota Department of Transportation 

• Dan Gullickson, Living Snow Fence Program Coordinator (TAP lead)
• Elizabeth Hobbs, GIS Technical Lead
• Ed Fleege, RWIS Project Administrator
• Lou Barrett, Transportation Program Supervisor
• Jim Klessig, Implementation Liaison
• Craig Schrader, MnROAD Coordinator

USDA - Natural Resource Conservation Service 

• Paul Flynn, Assistant State Conservationist for Technology 

Data and Software Support

Minnesota State Climatology Office and University of Minnesota, Department of Soil, Water, and Climate 

• Jim Zandlo, State Climatologist
• Greg Spoden, Assistant State Climatologist
• Pete Boulay, Assistant State Climatologist
• Dave Ruschy, Assistant Scientist

Please note: The lists above reflect individuals' affiliations and job titles as of the study's original publication in 2002.

Tabler, Ronald D., "Design Guidelines for Control of Blowing and Drifting Snow", SHRP-H-381, Strategic Highway Research Program, National Research Council, Washington, D.C., 1994.